Comparing Extreme Heat Flux Events for Wavenumber 1 and 2

Many studies of stratosphere-troposphere coupling have focused on composites based on the state of the polar stratospheric vortex (e.g. Charlton and Polvani (2007); Garnkelet al. (2010); Mitchell et al. (2013)). A smaller number of studies (Polvani and Waugh, 2004; Dunn-Sigouin and Shaw, 2015; Watt-Meyer and Kushner, 2015) have instead centered their attention on the flux into the stratosphere. We add to the latter class of studies by comparing composites of the largest positive and negative anomalous heat fluxes at 100 hPa for wavenumbers 1 and 2. Using MERRA 1 reanalysis data, we show that longer averaging times generally lead to fewer extreme events while the correlation to wave amplitudes of geopotential height and the zonal mean wind in the stratosphere becomes more robust. Integrating over the polar cap, wefind that the 100 hPa heat flux is well correlated with the flux at all levels in the stratosphere. However, the connection to tropospheric heat fluxes is limited to positive heat flux anomalies. A more detailed look at the longitudinal and temporal structure of the positive heat flux events allows us to identify statistically signicant anomalies in the geopotential heightfields associated with the anomalous fluxes. In the troposphere, wave-1 events are dominated by negative height anomalies over the Aleutian and positive values across most of the Atlantic. Wave-2 events are characterized by a more complicated NAM-like structure over the Atlantic and positive heights over the Aleutian. These results are compared to the results of Martius et al. (2009) and Colucci and Kelleher (2015) on the connection of blockings to stratospheric dynamics. Finally, we look at proles of planetary wave amplitudes and phases over the life cycle of the events. We find remarkably low variance in the phase as a function of height for both wave 1 and 2. This result is not sensitive to removing the climatological height field, further supporting the idea of linear interference as discussed by Watt-Meyer and Kushner (2015).